Third we performed correlation analysis between the results

Third, we performed correlation analysis between the results obtained using DMR agonist assay in HT-29 and Tango assay in the engineered U2OS cells. First, we compared their relative efficacy, the maximal responses after normalized to the maximal responses of zaprinast obtained using both DMR and Tango assays. Results showed that TTAC analogs displayed assay readout-dependent efficacy (a). These compounds can be classified to three categories using zaprinast as the reference compound. The first group includes compounds , , and that gave rise to comparable efficacy between the two assay readouts. The second group includes compounds such as , and that exhibited higher efficacy to cause β-arrestin translocation in the engineered Hydroxyfasudil than those to result in DMR in the native HT-29 cells. The third group includes compounds such as and which exhibited higher efficacy to trigger DMR than those to induce Tango signal. Compounds and exhibited the highest efficacy to cause β-arrestin translocation. Comparing their relative potency obtained using the two distinct assays showed that Tango assays generally gave rise to a right-shifted potency, compared to DMR assay results (b). Using YE210 as the reference, we found that most TTAC analogs led to a less right-shifted potency obtained using Tango assay. Compound displayed a moderate potency but the least right-shifted potency, compared to YE210. Compound was the most potent GPR35 agonist among all TTAC analogs tested. Finally, we examined the ability of compound to cause the internalization of endogenous GPR35 in HT-29 cells using a previously reported protocol. Results showed that similar to zaprinast at 10μM, compound at 12.5μM caused significant internalization of GPR35 in the native cell lines (). Together with DMR and Tango assays, this result further confirmed that compound is a potent GPR35 agonist. In summary, we synthesized a series of TTAC analogs and found that only compound exhibited higher potency than YE210. However, a couple of analogs including and displays higher efficacy than zaprinast to cause β-arrestin translocation. Further, correlation analysis suggests that many ligands displayed biased agonism. Future exploration of SAR may lead to the discovery of TTAC analogs with further improved potency, and more importantly, compounds having β-arrestin biased agonism for GPR35. Considering the importance of β-arrestin in regulating a diverse array of cell signaling pathway downstream GPCRs, such a class of GPR35 agonists may open possibility to further elucidate the biology and physiology of GPR35.
Introduction GPR35 is a Class A (rhodopsin-like) G protein-coupled receptor identified in 1998 [1]. The human GPR35 gene encodes a protein of 309 amino acids. GPR35 is expressed in various mammalian tissues, such as the gastrointestinal tissues, lymphoid tissues and the central and peripheral nervous tissues [1], [2], [3], [4]. Several investigators have reported GPR35 to be involved in the development of gastric cancer [4], the regulation of neuronal excitability and synaptic release [5], nociception [3], the pathogenesis of brachydactyly-mental retardation syndrome [6] and the regulation of blood pressure [7], although details are yet to be clarified. The endogenous ligand for GPR35 has long remained to be identified. Recently, Wang et al. [2] demonstrated that kynurenic acid, a tryptophan metabolite, is a possible endogenous ligand for GPR35. Kynurenic acid elicited several cellular responses in HEK293 cells and CHO cells expressing GPR35 [2]. Nevertheless, it remains unclear whether GPR35 is a specific receptor for kynurenic acid. Of note, relatively high concentrations of kynurenic acid (above 10μM) are necessary for eliciting apparent cellular responses [2], [3], [5], [8]. It seems curious that a G protein-coupled receptor would require such high concentrations of endogenous ligand for its activation. Thus, there may be other naturally occurring ligands for GPR35.